Skip to content

Commit

Permalink
Merged some files, changed names, updated code. hpr and hpr_mat have …
Browse files Browse the repository at this point in the history 
…a difference that has to be investigated. Fix coming soon
  • Loading branch information
George Stavrinos committed Jul 7, 2015
1 parent eec0173 commit 40e3e0d
Show file tree
Hide file tree
Showing 6 changed files with 171 additions and 943 deletions.
34 changes: 9 additions & 25 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ Real time recongition of humans through laser scans
python merge_train.py video/

7)Test online with the previously created classifier by running
python hpr_with_metrics.py video/Gaussian_NB_classifier_merged.p video/PCA_object.p scan 40 10
python hpr.py video/Gaussian_NB_classifier_merged.p video/PCA_object.p scan 40 10 0

8)In another terminal
cd
Expand Down Expand Up @@ -89,48 +89,32 @@ Runs the human pattern recognition (Naive Bayes) classifier.

Command line use:

python hpr.py <classifier_object_path> <pca_objec_path> <laserscan_topic> <timewindow_in_frames> <maximum_scan_range>
python hpr.py <classifier_object_path> <pca_objec_path> <laserscan_topic> <timewindow_in_frames> <maximum_scan_range> <0_or_1_for_metrics>

#V)hpr_with_metrics.py

Runs the human pattern recognition (Naive Bayes) classifier and generates a classification_results.mat file that contains its results.


Command line use :

python hpr_with_metrics.py <classifier_object_path> <pca_objec_path> <laserscan_topic> <timewindow_in_frames> <maximum_scan_range>

#VI)annotate_for_metrics.py
#V)generate_metrics.py

Runs the mat file that the classifier was tested on, and lets the user annotate the same clusters as the classifier in order to generate
some basic metrics (Precision, Recall, Accuracy).


Command line use :

python annotate_for_metrics.py </path/to/classification_results.mat>
python generate_metrics.py </path/to/classification_results.mat>

#VII)offline_test.py :
#VI)hpr_mat.py :
Runs the specified classifier on a .mat file instead of a topic.

Command line use :

python offline_test.py <.mat_file_path> <classifier_path> <pca_object_path> <timewindow> <frames for walls>

#VIII)offline_test_with_metrics.py :
Runs the specified classifier on a .mat file instead of a topic, and generates a classification_results.mat file that contains its results.

Command line use :

python offline_test_with_metrics.py <.mat_file_path> <classifier_path> <pca_object_path> <timewindow> <frames for walls>
python hpr_mat.py <.mat_file_path> <classifier_path> <pca_object_path> <timewindow> <frames for walls> <0_or_1_for_metrics>

#IX)Library files
#VII)Library files
gridfit.py, myhog.py and mytools.py are libraries used to run the code. They were pre-written by others or ported to python.

#X)online_calibration_node.py, optimal_test.py
#VIII)online_calibration_node.py, optimal_test.py
These two python files are not used anymore. They are still here just for legacy reasons and in case something new comes to our mind. You can freely ignore them.

#XI)ideal_data.mat
#IX)ideal_data.mat
This .mat file contains the basic information regarding the laser scans. It is not used by any of our currently used scripts, but it's still here for the sake of completeness
and future reference.

Expand Down
0 annotate_for_metrics.py → generate_metrics.py
View file
File renamed without changes.
177 changes: 118 additions & 59 deletions hpr.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -29,12 +29,17 @@
w_index=1
limit=3
scan_active = True
classification_array = []
scan_received = 0
plt.ion()
class_path = ''
pca_path = ''
pca_obj = PCA()
annotation_file = ''
first_time = True
first_time_ranges = True
sub_topic = 'scan'
metrics = 0

#temp2 = np.zeros((1, 36))

Expand All @@ -51,54 +56,80 @@ def RepresentsFloat(s):
return True
except ValueError:
return False

'''
def Calculate_Metrics(annotated_data):
global classification_array
pos = 1
true_pos = 0.0
true_neg = 0.0
false_pos = 0.0
false_neg = 0.0
neg = 0
#print len(annotated_data)
classification_array = np.array(classification_array)
#print len(classification_array)
for i in range(len(classification_array)):
if annotated_data[i]==1:
if classification_array[i]==1:
true_pos += 1.0
else:
false_neg += 1.0
else:
if classification_array[i]==1:
false_pos += 1.0
else:
true_neg += 1.0
precision = true_pos/(true_pos + false_pos)
recall = true_pos/(true_pos + false_neg)
accuracy = (true_pos + true_neg)/(true_pos + false_pos + true_neg + false_neg)
print "Precision : {0}".format(precision)
print "Recall : {0}".format(recall)
print "Accuracy : {0}".format(accuracy)
input ("Press any key to exit")
return
'''
def laser_listener():
#ADDITIONS command line inputs klp

print "###################################"
print "For non interactive input run as follows : "
print "python hpr.py <classifier_object_path> <pca_objec_path> <laserscan_topic> <timewindow_in_frames> <maximum_scan_range>"
print "##################################"
#ADDITIONS

global class_path, pca_path, sub_topic, timewindow, range_limit
global annotations
global gaussian, pca_obj
global timewindow, range_limit, annotated_data, classification_array

#ADDITIONS command line inputs klp
if len(sys.argv)>=3:
if not len(sys.argv) == 7:
print "###################################"
print "For non interactive input run as follows : "
print "python hpr.py <classifier_object_path> <pca_objec_path> <laserscan_topic> <timewindow_in_frames> <maximum_scan_range> <0_or_1_for_metrics>"
print "###################################"
exit()
else:
class_path = str(sys.argv[1])
pca_path = str(sys.argv[2])
print class_path
print pca_path

if not os.path.isfile(class_path):
while True :
try:
class_path=raw_input('Enter classifier object file path: ')
if os.path.isfile(class_path):
break
else:
print 'File does not exist! Try again!'
except SyntaxError:
print 'Try again'
print "Classifier File : {0}".format(class_path)

if not os.path.isfile(pca_path):
while True :
try:
pca_path=raw_input('Enter pca object file path: ')
if os.path.isfile(pca_path):
break
else:
print 'File does not exist! Try again!'
except SyntaxError:
print 'Try again'
print "File : {0}".format(pca_path)
#ADDITIONS command line inputs klp
global timewindow, range_limit
rospy.init_node('laser_listener', anonymous=True)
#ADDITIONS command line inputs klp
if len(sys.argv)==6:
if not os.path.isfile(class_path):
while True :
try:
class_path=raw_input('Enter classifier object file path: ')
if os.path.isfile(class_path):
break
else:
print 'File does not exist! Try again!'
except SyntaxError:
print 'Try again'
print "Classifier File : {0}".format(class_path)

if not os.path.isfile(pca_path):
while True :
try:
pca_path=raw_input('Enter pca object file path: ')
if os.path.isfile(pca_path):
break
else:
print 'File does not exist! Try again!'
except SyntaxError:
print 'Try again'
print "File : {0}".format(pca_path)

rospy.init_node('laser_listener', anonymous=True)
#ADDITIONS command line inputs klp
scan_topic = str(sys.argv[3])
timewindow = int(sys.argv[4])
while not RepresentsInt(timewindow):
Expand All @@ -110,21 +141,14 @@ def laser_listener():
range_limit = float(sys.argv[5])
while not (RepresentsInt(range_limit) or RepresentsFloat(range_limit)):
range_limit=input('Set maximum scan range in m: ')
if RepresentsInt(timewindow):
break
else:
print 'Try again'
else:
sub_topic = "scan"
while True:
timewindow=input('Set timewindow in frames: ')
if RepresentsInt(timewindow):
if RepresentsInt(range_limit):
break
else:
print 'Try again'
while True:
range_limit=input('Set maximum scan range: ')
if RepresentsInt(range_limit) or RepresentsFloat(range_limit):
metrics = int(sys.argv[6])
while not (RepresentsInt(metrics)):
metrics=input('Set maximum scan range in m: ')
if RepresentsInt(metrics):
break
else:
print 'Try again'
Expand All @@ -142,6 +166,27 @@ def laser_listener():
gaussian, pca_obj = loadfiles()
while not rospy.is_shutdown():
rospy.spin()
#we come here when Ctrl+C is pressed, so we can save!
if metrics == 1:
b={}
b['timewindow']=timewindow
b['range_limit']=range_limit
b['angle_increment']=angle_increment
#b['scan_time']=scan_time
b['angle_min']=angle_min
b['angle_max']=angle_max
b['intensities']=intensities
b['wall']=wall
print b['wall']
b['annotations']=annotations
b['ranges']=ranges_
try:
os.remove('classification_results.mat')
except OSError:
pass
sio.savemat('classification_results',b);
print "D O N E !"
#Calculate_Metrics(annotated_data)
#sys.exit()

def online_test(laser_data):
Expand Down Expand Up @@ -201,11 +246,12 @@ def online_test(laser_data):
ranges = ranges[filter]
theta = phi[filter]

#if first_time_ranges:
# ranges_= np.array(laser_data.ranges)[sampling]
# first_time_ranges = False
# else:
# ranges_ = np.vstack((ranges_, np.array(laser_data.ranges)[sampling]))
if metrics == 1:
if first_time_ranges:
ranges_= np.array(laser_data.ranges)[sampling]
first_time_ranges = False
else:
ranges_ = np.vstack((ranges_, np.array(laser_data.ranges)[sampling]))

if (len(ranges)>3): #each scan should consist of at least 3 points to be valid
#print "-------------- 6"
Expand All @@ -227,6 +273,7 @@ def online_test(laser_data):
z=- z_scale
z = z + z_scale
fr_index=fr_index+1
#classification_array = np.array(classification_array)


def pol2cart(r,theta,zed):
Expand Down Expand Up @@ -334,7 +381,8 @@ def scatter_all(xi,yi,zi,cluster_labels):

def update_plots(flag,hogs,xi,yi,zi,cluster_labels,vcl):

global kat, fig1, ax, wall_cart, gaussian, pca_obj
global kat, fig1, ax, wall_cart, gaussian, classification_array, pca_obj
global annotations, first_time
temp = []
#temp2 = np.empty(36) #Currently removed this way of calculating the zscore with temp2 because after an update on python it stopped working

Expand Down Expand Up @@ -373,15 +421,26 @@ def update_plots(flag,hogs,xi,yi,zi,cluster_labels,vcl):
[x,y,zed] = [xi[filter] , yi[filter] , zi[filter]]

if results[cnt]==1:
#classification_array.append(1)
kat.scatter(x,y,s=20, c='r')
ax.scatter(x,y, zed, 'z', 30, c='r') #human
fig1.add_axes(ax)
else:
#classification_array.append(0)
kat.scatter(x,y,s=20, c='b')
ax.scatter(x,y, zed, 'z', 30, c='b') #object
fig1.add_axes(ax)
cnt=cnt+1
plt.pause(0.0001)
if metrics == 1:
if first_time:
annotations = np.array(results)
first_time = False
else:
annotations=np.hstack((annotations,np.array(results)))
#b={}
#b['annotations']=classification_array
#sio.savemat('classification_results',b);



Expand Down
Loading

0 comments on commit 40e3e0d

Please sign in to comment.